Mean Residence Time of Plutonium in the Troposphere - American

Mean Residence Time of Plutonium in the Troposphere. Robert W. Holloway" and Davld W. Hayes. E. I. du Pont de Nemours & Co., Savannah River Laboratory...
0 downloads 0 Views 352KB Size
Environ. Sci. Technol. 1982, 16, 127-129

Mean Residence Time of Plutonium in the Troposphere Robert W. Holloway" and Davld W. Hayes E. I. du Pont de Nemours & Co., Savannah River Laboratory, Aiken, South Carolina 29808

Plutonium is injected into the stratosphere by large atmospheric thermonuclear explosions. When aerosols bearing plutonium reenter the troposphere, they are removed by such processes as rainfall and sedimentation. In this work, the mean tropospheric residence time of stratospherically injected plutonium is calculated. A mean residence time of 71 days was obtained for the Northern Hemisphere. Calculated residence times showed little variation by geographic location or for different years.

H

Introduction The mean residence time of aerosols in the troposphere is a quantity of considerable importance in relation to atmospheric pollutants, especially for those pollutants which are attached to the aerosol particles. Much effort has gone into the measurement of the aerosol residence time without producing an answer that is generally accepted. Some of the more common methods for estimating residence time use natural radioactive tracers such as 210pb, 210Bi,and 210Po(1-4). Other estimates have been made by using cosmic-ray-producednuclides and nuclear weapon debris. The apparent residence times calculated by these methods are often discordant, with the estimates ranging from 4 to 65 days (1-6). Since 1952, the stratosphere has had an inventory of artificial radioactivity more or less continuously. An estimate of aerosol residence time in the troposphere must then include the effect of the addition of radioactive particles from the stratosphere. Since the transfer process from the stratosphere has not been completely understood, natural radioactive tracers have been favored as a method of estimating the mean residence time of aerosols in the troposphere. The natural radioactive tracers used for this purpose are for the most part confined to the troposphere, and corrections for stratospheric input are not necessary. During the last 10 years, more information has accumlated concerning the transfer of material from the stratosphere to the troposphere. It now seems possible to estimate the mean tropospheric residence time of aerosols by using artificial radioactive tracers. In this work, an estimate is made of the mean residence time of plutonium, and the results are compared with the residence times obtained by other methods. Results and Discussion The calculated residence times shown in Table I are derived from 2399240Pu monthly concentrations in surface air as reported by the Environmental Measurements Laboratory (7). The data are combined into bimonthly averages of concentrations to reduce short-term fluctuations which are not of interest here. Residence times were calculated only for those periods which were free of lowyield atmospheric nuclear explosions, since these explosions would interfere with the removal rate determination. A June 1974 thermonuclear test by China injected debris into the stratosphere, but very little was detected in the troposphere during the summer and fall of 1974 (8). It *Address correspondence to this author at the following address: Environmental Protection Agency, P.O. Box 15027, Las Vegas, NV 89114. 0013-936X/82/0916-0127$01.25/0

Table I. Mean Residence Time of Plutonium in the Troposphere 1971a Mauna Lao, HI Moosonee, Ontario New York City Bimini, Bahamas Salt Lake City, UT Miami, F L av

65

r , day 1974a

89

65 79 81

66

C

85

69 58 12t12

82 C

77t8

1975b 61 44

66 C C

96 6 8 t 21

a Based on monthly concentrations whose standard deviBased on monthly ation of counting was less than 20%. concentrations containing one o r more months where the standard deviation was 20-100% and the remaining months less than 20%. Incomplete data.

therefore seemed reasonable to use that year for residence time calculations. The residence time of plutonium is derived from the rate of decrease of the tropospheric air concentrations during the summer and fall months. The initial concentration is taken to be the average concentration for May and June. This is often the peak concentration during the year, although the spring peak sometimes occurs during March or April. Assuming that the amount of debris which enters the troposphere from above during the months of July, August, September, and October is insignificant, one can easily obtain the troposphere residence time from the rapidly decreasing concentrations during the summer and fall. There is considerable evidence which indicates that the transfer of debris from the stratosphere is insignificant during July through October in comparison to the large amounts of debris which enters the troposphere during the spring peak. For example, Krey and Krajewski (9) concluded that the debris injected into the lower stratosphere in June 1967 by a 3-megaton Chinese explosion did not begin to leave the stratosphere until November of that year. Bauer et al. (10) concluded that material injected into the stratosphere in the summer does not leave until the following winter. More recently, Holloway (11)noted that only trace amounts of fallout were deposited in the summer and fall of 1973 in Arkansas even though a substantial inventory of debris existed in the lower stratosphere. The fact that fission product concentrations often decrease by 1order of magnitude in the months following the spring peak is also evidence that there is relatively little debris entering the troposphere during the summer and fall. The November-December concentrations are not used in the residence time calculations presented here since there is some evidence suggesting that stratospheric debris may begin to enter the troposphere during November and December (10). The time (2') required for the tropospheric concentrations of the tracer to be reduced by half is given by the equation T = 0.693t/ln (Q1/Q2) (1) where Q1 is the concentration of the initial inventory as

0 1982 American Chemical Society

Environ. Sci. Technol., Vol. 16, No. 2, 1982

127

6

- 5 0 .-c

e

c

C

;4 0 0 3

a $

3

OF 0

6

2

1

C

Figure 1. Seasonal variations of the n s , 2 ~concentrations u in surface air at Miami, FL, in 1971.

measured by the May-June concentration, Qz is the concentration at some later time, and t is the time interval in days between Q1 and Q2. The time (7') required to remove half of the debris is less than the mean residence time (T), and they are related by the equation T

= T/0.693

(2)

The mean residence time then can be expressed as 7

= t/ln

(Qi/Q2)

(3)

At a particular location, the concentration of 239~240Pu in the atmosphere decreases approximately exponentially

during July through October in the Northern Hemisphere. However, the bimonthly concentrations seldom fit precisely on a line corresponding to a specific residence time. To obtain the best estimate of residence time at each location, we converted the bimonthly concentrations for May through October to natural log form, and a line was fitted to the three points by the least-squaresmethod. The mean residence time was calculated by using eq 3 and Q values obtained from the fitted line. The calculated residence times are given in Table I. Figure 1 shows the typical seasonal variations in plutonium concentrations at a single location. Note that the bimonthly concentrations for May through October deviate slightly from a straight line. These deviations are typical for the concentrations measured at a single location and probably result from meteorological factors rather than analytical error. Transient meteorological conditions on a time scale of a few weeks could cause variations in the removal rate over fairly large regions. The result would be variations of the measured concentrations about the line corresponding to the best single residence time. In order to reduce the influence of regional effects, it seemed appropriate to average the bimonthly concentrations of the widely dispersed sampling locations. The results are shown in Figure 2. The summer and fall bimonthly concentrations fit the line corresponding to a single residence time remarkably well. The air concentration measurements for the locations in Table I were used to obtain the averages plotted in Figure 2. The resulting residence times (7) in Figure 2 can be considered as typical of the Northern Hemisphere since they are derived from air concentrations at widely dispersed locations. The average of the three residence times shown in Figure 2 is 71 days. This is higher than most previous calculations of residence time, but it is in good agreement with the results of Marenco and Fontan (21, who calculated a global tropospheric residence time of 65 days based on the 210Po/210Pbratio. The longer residence times of 65 and 71. days do not necessarily conflict with estimates in the 30-40-day range. Marenco and Fontan's 65-day result was obtained from the z10Po/210Pb ratio throughout the entire troposphere (2). The 35-day estimate of Nevissi et al. ( 4 ) was based on the same ratio in rain samples. The 71-day

Flgure 2. Average blmonthly concentrations of 239*240Pu for the Northern Hemisphere. The locations used to obtain the average concentrations are those listed in Table I. 128

Envlron. Sci. Technol., Vol. 16, No. 2, 1982

residence time obtained in this work is derived from particles that have been transported downward from the stratosphere to near the earth’s surface. It is reasonable to expect some variations in calculated residence time if the particles being sampled have had quite different origins. In the case of 210Pband 210Po,both are formed in the troposphere from the decay of 222Rn. An additional source of 21”Po from volcanic emission has been suggested recently by Lambert et al. (12).Their results indicate that volcanoes may be a significant source of 210Pocompared to the decay of atmospheric 222Rn. Since their estimate of volcanic zlOPois an order of magnitude gstimate, the extent to which the tropospheric 210Po/210Pbratio is influenced by volcanic emissions is not clear. Moore et al. (13) favor a residence time of only 4-6 days in the troposphere and suggest that soil particles and plant exudates are the major sources of 210Poin the atmosphere. The particle size distribution of phtonium-bearing aerosols may also be slightly different from the size distribution of particles containing natural radioactive tracers. This might lead to different residence times for each type of aerosol. Fission products in the stratosphere are attached to particles whose average diameter is approximately 0.2 pm (14). After these particles enter the troposphere, a definite shift toward larger mean diameter has been observed. For example, Gasiev et al. (15)found that only 3% of the /3 activity at ground level was associated with particles with a diameter of less than 0.3 pm, while at an altitude of 5-6 km about 48% of the p activity was associated with pa?ticles of that size fraction. The shift toward larger particles in the troposphere is thought to b6 caused by coagulation of the radioactive particles with natural aerosols. Presumably, natural radioactive tracers such as 210Pband 210Poundergo the same process, though the original attachment of 210Pband 210Powould be to particles of less than 0.3-pm diameter (16).It should be noted that Moore et al. (17)have found cases where the shift to lalger particle sizes did not occur for Y3r as measured in surface air. Because of the stratospheric source of the plutonium-bearing particles and possible differences in their size distributions compared to other tracers, the residence time of 71 days may be slightly more than the true residence time of natural aerosols. However, our results tend to support these estimates which range from 30 to 65 days rather than the much shorter residence time estimates of 5 to 10 days. There is also the possibility that the 71-day residence time estimated from plutonium-bearing particleg is valid for natural aerosols in general. The good agreement with the 65-day estimate of Marenco and Fontan (2)

tends to support this interpretation.

Conclusions The removal rate of plutonium-bearingaerosols from the troposphere can be calculated from the decrease of the tropospheric inventory present during the spring. The removal rate corresponds to a mean residence time of 71 days, This is a Northern Hemispheric residence time and only slight variability was observed at widely dispersed locations. The mean residence time for plutonium should be similar to that of natural aerosols, although not necessarily identical. Literature Cited (1) Gavini, M. B.; Beck, J. N.; Kuroda, P. K. J. Geophys. Res. 1974, 79, 4447-52. ( 2 ) Marenco, A.; Fontan, J. Tellus 1972, 24, 38-46. (3) Moore, H. E.; Poet, S. E.; Martell, E. A. J. Geophys. Res. 1973, 7& 7065-74. (4) Nevissi, A.; Beck, J. N.; Kuroda, P. K. Health Phys. 1974, 27, 181-8.

( 5 ) Stewart, N. G.;Crooks, R. N.; Fisher, E. M. U.K. At. Energy Res. Establ. 1955, Rep. AERE HPJR 1701. (6) Martef, E: A,; Moore, H. E . pech. Atmos. 1974, 903-10. (7) Environ. Meas. Lab. Envirdn. Q (U.S. Dep. Energy) 1980, EML-370, Appendix, pp 153-6. (8) Holloway, a. W. Ph.D. Thesis, Department of Chemistry, University of Arkansas, Fayetteville, AR, 1974. (9) Krey, P. W.; Krajewaski,’B.’J. Geaphy. Res. 1970, 75, 2901-8. (10) Bauer, E.; Oliver, R. C.; Waslkiwskj, W. J. Geophys. Res. 1978,83, ’4019-25. (11) Holloway, R. W. Environ. Sei. Technol. 1979,13, 1488-90. (12) Lambert,G.; Buisson, A.; Sapak, J.; Ardouin, B. J. Geophys. Res. 1979, 84, 6980-6. (13) Moore, H. E.; Martell, E. A.; Poet, S. E. Enuiron. Sci Technol. 1976,10, 586-91. (14) Martell, E. A. Tellus 1966, 18, 486-98. (15) Gasiev, Y. I.; Malalihov, S. G.; Nazarov, L. E.; Silantiev, A. N. Tellus 1966, 18, 474-84. (16) Junge, C. E. “Air Ch*istry and Radioactivity”; Academic Press; New York, 1964. (17) Moore, H. E.; Poet, S. E.; Martell, E. A. In “Natural Ra~

diation Environment” besell, T.’F., Lowder, w. M., Eds.; NTIS: Springfield, VA, 1980; Vol. 111, pp 415-29.

Received for review July 21,1981. Revised manuscript received September 24,1981. Accepted November 12,1981. This paper was prepared in connection with work under Contract No. DEAC09-76SR00001 with the US Depqrtment of Energy.

Environ. Scl. Technol., Vol. 16, No. 2, 1982

129